Cross-study single-cell RNA analysis

Integrate multiple scRNA-seq datasets across studies

Building cell atlases and assessing reproducibility of single-cell findings across studies with detection-aware quality filtering.

• · ENCODE scRNA-seq datasets from multiple studies/donors
• · Sample metadata (tissue, cell type, platform)

• · Batch-corrected integrated matrix (genes × cells)
• · Reproducibility-scored cell type assignments
• · Cross-study quality metrics (TIN scores, detection limits)

• · Seurat 3 / Harmony / LIGER
• · Python (scanpy) or R (Seurat)
• · ENCODE database API

Multiple scRNA-seq count matrices in standard format, knowledge of cell type biology

Severe batch effects persist after correction, detection-limit artifacts mask true heterogeneity, cross-contamination from ambient RNA, dropout prevents marker detection

• · Mawla et al. 2019 meta-analysis framework
• · Cross-study reproducibility focus
• · TIN-based quality assessment
• · Stuart et al. 2019 (8400+ citations), Luecken & Theis 2019 (1631 citations)